Measuring the mathematics abilities of students with special education needs through a computer-based multilevel adaptive test

Papers in this third volume deals with the Covid-19 pandemic which is having an enormous impact on education systems worldwide. Policy makers, teachers, school managers, parents and students have been called to the reinvent their way of ‘doing school’. At the same time, the governance of the education system and schools’ organizations have been exposed to unprecedented tensions. The pandemic emergency has been a lens revealing intersections and structural tensions among various level and actors of the education system, but also allowing opportunities of changes thanks to the exogenous shock. At the same time, it must be considered that the emergency is interacting on pre-existing inequalities and contradictions. Higher education systems have been affected too: in constant evolution due to constant transformations of society and changed functions of knowledge, universities have undergone a structural change along with pandemic times. Simultaneously, the growing relevance of knowledge for the economic development of the capitalistic system has profoundly affected higher education systems, characterized by the neo-liberal approach which his subject of increasing critical analysis.

In this paper we present results on the estimation of mathematics skills of Italian grade 10 students with special educational needs. The estimation was carried out by means of a com-puter based multilevel adaptive test 1 - 3 - 3 (MST) constructed within the framework of a doctoral research in Social, Developmental and Educational Research Psychology. An item bank was used to construct the MST test. Verification of the unidimensionality of the bank was done through exploratory factor analyses (EFA) and analysis of model fit indices with Rasch's model (Rasch, 1960) being calibrated. Verification of the best measurement properties of the test compared to a linear test was performed by comparing information functions. It is shown that this type of test is able to provide estimates of the ability of students with a con-siderable level of accuracy, even at the extremes of the range of abilities and not only in a neighborhood of the mean value. In particular it was observed that the test is also suitable for students with special educational needs. Such students are distributed across all ability levels, although the majority of them are in the low ability range.